HiringGPT: Empowering HR Through Generative AI While Keeping the Human Touch
In the always changing landscape of Human Resources (HR), the integration of technology has been a double-edged sword. On one hand, it promises efficiency and scalability; on the other, there’s a pervasive fear of losing the invaluable human factor that is so important when making decisions that directly impacts people’s lives.
However, what if there’s a way to leverage cutting-edge technology not to replace, but to augment the human element in HR? Our approach to harnessing generative AI in the HR process illustrates this possibility, augmenting the capabilities of HR professionals while ensuring that personal touch remains at the center of hiring.
From Job Description to Candidate Shortlisting
The process begins as it usually does, with posting a job description (JD). Traditionally, drafting a JD is the step in which a company both announces what role they are trying to fill, as well as the skills needed to do so, and also nice-to-haves and other factors that might impact the decision to hire someone. As for the candidate’s side, the JD tells them whether or not to even bother applying.
However with our approach it can do more, it also signals the LLM to spring into action. Based on the JD, the model suggests an initial set of criteria to look for in candidates — ranging from technology stacks and work experience to more nuanced qualifications, all tailored to the specific position. This makes the initial process of deciding the criteria to look for faster, but also these decisions are not locked in place, as the goal is to have the human lead in every step of the process, with the AI taking an auxiliary role.
Customizing Model Suggestions
Crucial to this process, is keeping the HR manager in the lead. The model’s suggestions serve as merely a starting point. From there, HR managers can adjust these criteria, add new ones, or refine existing parameters to fit the unique needs of the position and the company culture. This could include specifics like eligibility to work in the country or visa support requirements, ensuring that the model’s next steps from data extraction to search is as close to the manager’s expectations as possible.
Extracting What You Need
Once the criteria are set, generative AI sifts through hundreds of applications, extracting relevant information and standardizing it into a uniform data model. This transformation is crucial; it converts diverse candidate information into a format that’s easy to navigate and compare. Everyone has their own way of filling out a resume, but the things the HR manager needs to know in order to make an informed decision is actually rather structured, so putting all this different information together in a way that conforms to what the hiring manager is looking for is pivotal for this process to work.
Finding the Perfect Match
With this standardized data at hand, HR managers can then utilize a dynamic filtering system. Want a candidate with Python experience, who has worked with transformer models and boasts over three years of experience in Natural Language Processing (NLP)? Just set your filters. This system drastically reduces the pool from hundreds to those who meet all specified criteria, making the evaluation process not just faster, but more effective, while remaining fair.
Feedback and Communication
The final, and perhaps most innovative, aspect of this approach is how to handle communication with candidates. For those who advance, the next steps are straightforward. However, for those who don’t, the model crafts personalized feedback based on the criteria they didn’t meet. This feedback is invaluable, offering candidates clear insights into their application’s shortcomings and reducing the frustration associated with opaque (or even missing) rejection notices.
Empowering HR, to the Benefit of Candidates
What sets this approach apart is not just the efficiency it introduces to the hiring process, but the capabilities it provides to hiring managers and the empathy it extends to candidates. By leveraging generative AI, we’re not sidelining the human element; we’re amplifying it. HR professionals can achieve more, with greater precision, without sacrificing the personal engagement that defines the best hiring practices. Meanwhile, candidates receive a more transparent and constructive application experience.
This use of generative AI in hiring can therefore serve as a sign of the potential of technology to enhance human capabilities, rather than diminish them. In a work environment and labor market where people are increasingly anxious about the future of their career, it is our responsibility to make sure that we are using the technology in the best possible way, from where the employer/employee relationship begins, the hiring process.
Keeping People in the Driver’s Seat, at Every Step of the Way
One of the big concerns when utilizing machine learning techniques, which has prompted both concerns from the general public and governmental bodies, is the matter of accountability, liability, and leaving decisions that impact the livelihood of many to a machine.
It is with this concern in mind that this approach was developed, the power of generative models are leveraged to turn large amounts of unstructured data into a structured format, however no scoring, sorting, evaluation, filtering, or anything of the sort is done by the algorithm. This ensures that it will always be a human in charge of the selection criteria, evaluation, and ultimately hiring.
What we ultimately achieve is thus a way to both maintain a commitment to strict ethical standards as well as boosting the productivity and expanding the capabilities of hiring managers across the board.
